The Government may own certain rights in the present invention pursuant to NIH grant numbers CA21927, CA38499 and CA39798.

The present invention relates to the use of recombin¬ ant DNA technology to clone and sequence genes which code for proteins exhibiting human B-cell growth factor activity and, more particularly, the use of these cloned genes in the production of such proteins.

Our basic knowledge of the mechanisms underlying immune regulation by the human body has substantially increased during the past decade. This increase in understanding has been due, in part, to knowledge gained with respect to a group of soluble factors secreted by lymphoid cells. In general, these soluble factors serve to mediate intercellular cooperation. For example, studies have indicated that many of the events making up

the immune system's response to antigenic challenge are mediated by certain soluble factors termed lymphokines and onokines, which are secreted by specific cells of lymphoid and monocytoid derivation. For a review of the actions of various soluble factors which have been identified see Maizel et al. (1984), "Biology of Disease," in Laboratory Investigation, 50; 369-377.

One of these lymphokines, designated T-cell growth factor (TCGF) or interleukin-2 (IL-2), has recently been reported by Gallo and his co-wakers at NIH to provide a mechanism for specifically targeting activated "killer" T-cells to human cancers. Moreover, the ability to grow some forms of neoplastic mature T-cells using TCGF, directly led to the isolation of human T-cell leukemia virus (HTLV) , instrumental in the development of an immunologic test for acquired immune deficiency syndrome (AID's). Much of the success of IL-2 as a potential anti-cancer agent can be attributed, in part, to the availability of IL-2 produced by recombinant DNA technology.

B cell growth factors have been identified which stimulate B-lymphocyte proliferation. It has been observed that BCGF's exhibit acivities that are roughly analogous to various activities exhibited by TCGF. Although no BCGF species has yet been isolated to sufficient purity and in sufficient quantity for confir¬ mation, it is thought that BCGF may play a similarly significant role in treating certain human diseases such as cancer.

One reason for our lack of knowledge concerning this protein(s) is that no recombinant clone is presently available for the large scale production of human BCGF— BCGF that would be free of interfering cytokine activity.

The availability of an anti-BCGF antibody would provide a reasonably reliable probe for the isolation of a recombinant-BCGF producing cellular clone. Unfortunately, although BCGF has been isolated on an analytical level from lectin-activated peripheral blood mononucleocytes (Maizel et al. (1982), Proc. Natl. Acad. Sci. U.S.A., 79: 5998), reseachers have been unable to produce such an anti-BCGF antibody.

Therefore, the art has heretofore been unable to circumvent those problems inherent in the preparation and selection of recombinant cellular, clones in the case of B-cell growth factors. Such problems are compounded where, as with molecules exhibiting BCGF activity, there exists no clone selection antibody. Therefore, for the foregoing reasons, a recombinant clone providing BCGF activity independent from other cytokine activity would represent an advance in medical science. Additionally, techniques for providing such a recombinant clone would represent a similar advance.

Accordingly, it is an object of the present invention to provide cellular clones which produce proteins which exhibit activity corresponding to human B-cel-1 growth factor.

It is a further object of the present invention to provide such a protein by recombinant DNA techniques.

It is a still further object of the present invention to provide techniques for the preparation, identification and isolation of cellular clones which produce a recombin¬ ant protein exhibiting human B-cell growth factor activity.

It is yet an additional object of the present inven¬ tion to provide recombinant isolation techniques which circumvent a requirement for an anti-BCGF antibodies as probes to identify such recombinant cellular clones.

Accordingly, in its most general and overall scope, the present invention discloses a DNA segment comprising a DNA sequence which codes for a protein exhibiting human B-cell growth factor activity. Such DNA segment may conveniently be provided in the form of a recombinant DNA vector which includes the DNA sequence. Recombinant DNA vectors made in accordance with the present invention and when introduced into an appropriate host, provide such host with the genetic information necessary for the production of an active recombinant human B-cell growth factor.

Thus, the present invention discloses techniques suitable for the construction of recombinant DNA vectors which contain a DNA sequence which codes for a protein exhibiting human BCGF activity. In addition, the present invention discloses a method for producing genetically engineered human BCGF utilizing recombinant DNA sequences which code for a human BCGF. These sequences will generally be referred to as a recombinant human BCGF gene.

Genetically engineered BCGF is produced by providing a cellular clone containing a recombinant DNA expression vector, wherein the expression vector contains a DNA sequence which codes for human BCGF, and culturing the cellular clone under suitable conditions to promote the transcription and translation of the recombinant human BCGF gene into a human BCGF. A recombinant DNA expression vector as used herein refers to a cloning vector, such as the pUCq vector described herein, which has been particu-

larly adapted for expressing the protein product coded for by the inserted recombinant DNA sequence.

A cell which expresses human BCGF is provided by first, preparing at least one first nucleic acid sequence which is complementary to at least one second nucleic acid sequence wherein the second nucleic acid sequence codes for human BCGF, followed by transforming an appropriate host cell with the first nucleic acid sequence so prepared. A cellular clone actively expressing human BCGF can then be selected from the group of cells so trans¬ formed.

Although genetically engineered BCGF may be provided in the above manner by providing the recombinant expres¬ sion vector directly, such protein may similarly be provided from recombinant DNA trasfer vectors, such as pBR322, and which DNA sequences include a code for human BCGF. In this manner, a recombinant expression vector containing the recombinant BCGF gene DNA may then be constructed using the recombinant BCGF gene DNA sequence contained within such recombinant transfer vector. A recombinant DNA transfer vector as used herein refers to a cloning vector which may be used in transferring cloned recombinant DNA sequences from one host to another and are particularly useful for preparing recombinant clone banks. Examples of transfer vectors include pBR322 and its numerous derivations. It should be noted that transfer vectors can be used for expressing a cloned gene, however such vectors are generally not as efficient as the expression vectors for such purposes.

Recombinant DNA vectors as used herein may be either recombinant DNA transfer vectors or recombinant DNA expression vectors. Such recombinant DNA vectors include but are not limited to plasmids (which are extrachrom-

asomally replicating circular DNA's), bacteriophage (such as the lambda phage), and human viruses such as SV-40 which have been adapted for both transfer and expression of recombinant DNA sequences (see, e.g., Okayama and Berg, infra. ) .

One method for preparing at least one nucleic acid sequence which is complementary to at least one second nucleic acid sequence which codes for human BCGF entails enzymatically copying total human messenger RNA (mRNA) , preferably RNA from activated T-lymphocytes, into complementary DNA (cDNA) sequences. In this manner, at least one of the cDNA sequences so obtained should contain the necessary genetic sequence information to code for a human BCGF mRNA. A recombinant clone which receives a cDNA complementary to a gene coding for human BCGF and which is actively producing human BCGF can then be selected using the selection techniques described herein.

The preferred enzyme for enzymatically copying human mRNA into complementary nucleic acid sequences is the enzyme reverse trascriptase. However, other nucleic acid polymerizing enzymes, such as DNA polymerase I, may be useful. Nucleic acid sequences capable of coding for a human BCGF may be prepared in other ways as well, includ¬ ing synthetic prepartion of the appropriate DNA sequence, using the nucleic acid sequence or amino acid sequence of Figure 5 as a guide. Synthetically prepared BCGF gene sequences should function identically with BCGF gene sequences which are enzymatically prepared.

Although for commercial preparation of recombinant BCGF it may be more appropriate to provide the recombinant BCGF cDNA in a bacterial host/vector system, it is never- theless possible to provide the cDNA in a eukaryotic

host/vector system using the techniques described or referred to herein.

BRIEF DESCRIPTION OF THE DRAWINGS

Fig. 1A exhibits the enrichment of BCGF specific mRNA by methyl mercuric hydroxide gels. Symbols represent activity for: ( ) BCGF; ( __ ) TCGF. The left sided ordinate represents thymidine incorporation in the BCGF assay; the right sided ordinate represents incorporation in the TCGF assay.

Fig. IB exhibits the translation products of hybrid selected BCGF mRNA's in Xenopus laevis. Results are represented in terms of 3H-Tdr incorporation in 12 x 103 long term cultured B cells. Symbols represent DNA from: ( ♦ ) pUC 9; ( > ) pARJ 43; ( a__ ) pARJ 45; ( __. ) pARJ 55; ( • ) pARJ 36.

Fig. 2A demonstrates the biological activity of recombinant BCGF. Results are represented in terms of

3 H-Tdr incorporation of a representative microtiter culture at 5% final concentration of each fraction. Symbols represent cell extract from: ( _» ) pUC 9; ( ___ ) pARJ 43; ( α ) pARJ 45.

Fig. 2B demonstrates specificity on anti-u activated B cells. Fractions from panel A were also assayed on anti-u activated B cells at 1.2% final concentration essentially as described for panel A, except incubation in the microtiter assay continued for 96 hr before addition ooff ³ HH--TTddrr aatt 772 hrs. Symbols represent: ( # ) pUC 9; ( A ) pARJ 43.

Fig. 3 is a representative restriction map of pARJ 43. A pR designates the location of the ampicillin

resistence gene. Lac Z indicates the location of the B- galactosidase structural gene.

Figs. 4A and B together display the DNA sequence of the BCGF insert in pARJ43 along with the corresponding derived amino acid sequence. The numbers refer to nucleotide sequence with respect to the first BCGF-coding nucleotide. Also displayed are the various endonuclease restriction sites.

In that Figures 4A and 4B together illustrate one continuous nucleotide and protein sequence, they will be referred to throughout the present disclosure as simply Figure 4. From the nucleotide numbering system shown in Figures 4A and 4B, it will be readily apparent to those of skill in the art that the leftmost nucleotides of Figure 4B should be aligned to the right to the rightmost nucleotides of Figure 4A to display the entire continuous sequence.

Human B-cell growth factor (BCGF) is a T-cell derived lymphokine under normal circumstances, that has been shown to be involved in the proliferation of activated B lymphocytes. Biochemical characterization of human BCGF has revealed that it is a heat-and protease-sensitive protein which exhibits an approximate molecular weight of 12,000 to 14,000 with a major isoelectric point of 6.5. Although originally described as a factor functionally specific to promote S phase entry of B cell, BCGF may also belong to a cascade of signals that are involved in the initial activation of cells of this lineage.

BCGF is predominately released by normal human T cell lymphocytes following antigen and/or lectin stimulation of such cells. Similarly, BCGF activity has been shown to be released from certain granular lymphocytes and "immorta-

lized" human B cells. Moreover, multiple BCGF species of different molecular weights have been detected in super- natants of T-cell hybridomes, malignant T cells and normal lectin activated T cells costimulated with PMA. The numbers and variations of BCGF's suggest the existence of multiple BCGF species. Unfortunately, in all of the foregoing cases, the molecules exhibiting BCGF activation activity are not produced in sufficient quantities to render the isolation of significant amounts of BCGF practical.

The present invention takes advantage of recombinant DNA technology to provide cellular clones which produce a protein having BCGF activity. For the purposes of illus- trating the present invention, protein molecules which exhibit BCGF activity will be referred to generally as BCGF. In general, the production of recominant human BCGF is achieved by, first, selective enrichment for BCGF- specific messenger RNA (mRNA) ; second, preparation of complementary DNA (cDNA) segments which correspond to the BCGF-enriched mRNA fraction and; thirdly, inserting the resultant BCGF-enriched cDNA's into an appropriate expression host/vector cloning system. Such DNA expression vectors are well known to those of skill in the art to be useful in both the cloning of recombinant gene segments and in transforming their hosts to produce the corresponding gene segment protein.

In a preferred embodiment of the present invention, the pUC-9 vector as described by Heidecker et al. (1983), Nucl. Acad. Res. , 11: 4891, is utilized to take advantage of insertion of the cDNA sequences into the B-galacto- sidase gene which it possesses. However, it is comtem- plated that full advantages of the present invention may be realized with other host/vector expression systems, for example, the pIN-III vectors as described by Mafsui et al.

(1983) in Experimental Manipulation of Gene Expression, edited by M. Inoye, or the SV-40 derived eukaryotic vector of Okayama and Berg (1983). Mol. Cell. Biol. , 3.: 280. The Okayama vector permits expression of proteins in selected eukaryotic cells. Numerous other appropriate host/vector systems are known to the art and are contem¬ plated to be within the scope of the present invention.

Following the preparation of representative BCGF- enriched cDNA-containing recombinant expression vectors, and the introduction of such recombinant vectors into an appropiate expression host, the problem still remains as to the selection of the appropriate BCGF-producing clones from the hundreds or thousands of recombinant clones which are produced. Typically, such a selection has involved the use of an antibody to the desired protein to antigen- ically identify those clones actively expressing the desired protein. However, such an antibody would tend to identify clones which produce both "active" and "inactive" forms of the desired recombinant proteins. That is, antigenic detection methods would also select for clones which include inactive reproductions of the desired proteins. For this reason, and the fact that the production of anti-BCGF antibodies has not been achievable, the development of alternative approaches was thus necessitated.

To clone the BCGF gene, advantage may be taken of certain observations. Human B-cell lines can be established which depend upon either mature BCGF (12-14

KD), or larger putative intracellular precursors, for

3 their proliferation (measurable by H-thymidine uptake into newly synthesized DNA) . Such cell lines provide a convenient system to sensitively assay either full length or fragmented BCGF molecules produced by recombinant methods. However, because it is a biological assay, it

selects for only those clones which produce a biologically active position. The utilization of the biological assay additionally provides for lymphokine specificity in selections for BCGF activity.

For the cloning described herein, full advantage of the present invention may be realized by preparing such BCGF dependent target B cell lines which are continuously selected for a demonstrated reactivity to BCGF in the essential absence of proliferative reactivity to other cytokines such as interleukin-1, interleukin-2, or gamma- interferon. Such "closely-selected" BCGF-target cells therefore provide for the identification of BCGF activity without identifying other cytokine activities.

b. BCGF Isolation

The proper selection of BCGF-dependent B cell lympho¬ cytes, described in section c. below, is therefore of particular importance in the development of a sufficiently BCGF-specific bioassay. Additionally, such cells are necessary to provide an assay that is sufficiently BCGF- sensitive when used in conjunction with mRNA-primed Xenopus oocyte translations. Thus, to provide sufficient- ly BCGF-dependent B cell lymphocytes, it is generally preferrable to select such cells in the presence of highly purified quantities of BCGF. The following techniques provide such purification.

To obtain human BCGF(s) from a primary source (i.e. a non-recombinant source), it is generally preferrable to isolate it from the culture supernatants (CM) of BCGF producer cells. The kinetics of the accumulation of BCGF activity in culture supernatants obtained from lectin- stimulated, peripheral blood mononuclear cells reveals a maximum yield from 48 to 72 hr of culture. To maximize

the presence of stable amounts of BCGF, it is generally necessary to add _0.25% human serum albumin during the initial generation phase of the CM.

The CM is then concentrated 20-fold using a Pellicon (Millipore) concentrator containing a 10,000 Dalton m.w. cut-off membrane. The concentrated material is subse¬ quently diafiltered by using 10 vol. of lOmM Tris, pH 8.0. After centrifugation at 10,000 x G for 10 min. the clarified sample is loaded onto a previously equilibrated

2 DEAE-Sepharose column at a rate of 5 to 10 ml/cm hr. The column is eluted with a linear gradient of 0 to 0.2 M sodium chloride in 10 mM Tris buffer, pH 8.0. Fractions of approximately 12 ml are collected and bioassayed for BCGF and TCGF activities as described below. BCGF activity elutes from the column at relatively low ionic strength (0.08M). However, the fractions possessing BCGF activity also generally contain other cytokines, incuding TCGF.

Prior techniques for isolating highly purified cytokines have generally required the use of reverse-phase high pressure liquid chromatography. However, due to the extremely low recovery (>1%) of BCGF activity from reverse-phase high pressure liquid chromatography (RP- HPLC), it is generaly preferrable to chromatograph DEAE- purified BCGF on hydroxylapatite columns. Thus, in a preferred method for the removal of TCGF activity, biologically active material from the DEAE-Sepharose column is concentrated 150-fold in a Millipore concentrator and dialyzed (using a Spectrapor 3 membrane, 3500 m.w. cutoff) against 10 mM sodium phosphate, pH 6.8. The concentrated sample (equivalent to 1 liter of crude CM) is then loaded onto a hydroxylapatite column (1-mm column with 30-ml bed volume) previously equilibrated with 10 mM sodium phosphate buffer. The column is subsequently

washed with 0.04 M sodium phosphate until absorbance at 280 nm is reduced' to base line. The nonbound material is collected, whereas bound material is eluted with 150 mM sodium phosphate, pH 6.8. The flow-through and bound eluants are bioassayed for both BCGF and TCGF activities as described below.

This column chromatographic step of purification provides relatively good recovery (>24% yield) of BCGF biologic activity and generally yield BCGF essentially free of other important contaminating cytokine activities.

The hydroxylapatite-purified BCGF activity may next be chromatographed on an analytical HPLC-DEAE anion- exchange column (Protein Pak 5PW; Waters Associates) equilibrated in 15 mM phosphate buffer, pH 8.0. :

In a typical experiment, ten to fifteen milligrams (0.5-ml samples) of the hydroxylapatite-purified, TCGF- free BCGF preparation was loaded onto a Waters Associate DEAE 5PW, Protein Pak anion-exchange column (7.5 mm x 7.5 cm) previously equilibrated in 15mM sodium phosphate buffer, pH 8.0, at a flow rate of 0.5 ml/min. Ater a 2- min isocratic run, a linear gradient of 0 to 0.2 M sodium chloride in the same buffer was applied at a flow rate of 0.5 ml/min over the next 30 min. by using a Waters automated gradient controller. The gradient was held at 0.2 M sodium chloride in phosphate buffer for the next 5 min. then was increased to 1 M sodium chloride for complete removal of bound proteins. The optical density of eluting proteins was monitored at 229 nm by a Pharmacia chart recorder at a speed of 3 in./hr.

The loosely bound BCGF bioactivity eluted at 0.07 to 0.10 M salt concentration. The recovery of the eluted BCGF bioactivity showed approximately 65% of the

bioactivity contained in 1% of the total protein content. The column was regenerated for repeat use after increasing the gradient concentration to 1 M salt to elute bound contaminating proteins and then equilibrating to the starting phosphate buffer containing no salt. A highly reproducible elution pattern permitted the collection of about 10 consecutive runs of BCGF within a narrow gradient concentration or salt without any further loss of bioactivity or excessive contamination of closely spaced protein peaks.

BCGF activity isolated in a manner as described above may be futher purified on HPLC size exclusion columns. In a typical experiment, a sample of reconstituted material (0.5 ml) obtained after affinity HPLC and lyophilization was loaded onto gel filtration columns (1-60 + 1-125 + I- 60; Waters Associates) in series that were previously equilibrated in 20 mM phosphate buffer, pH 7.0, at a flow rate of 0.5 ml/min. A molecular weight calibration was performed by using bovine serum albumin (67,000 dalton) and myoglobin, (18,000 dalton) as standards, the BCGF bioactivity eluted as a sharp peak at a m.w. of 12,000 to 14,000, as determined by calibration of the column with the m.w. standards. - These fractions displayed BCGF activity when assayed with either anti-u-activated human B cells or BCGF-dependent long-term B cells lines. The recovery of the biologic ativity from the gel filtration was >35% of the applied BCGF activity.

Previous attempts at purifying human BCGF from peripheral blood T cell sources that include the co- eluting lymphokine TCGF have been successful only at an analytical level. In these prior procedures, a series of gel filtration columns (Biogel P30 and P100) were employed to obtain 12,000 to 14,000 dalton BCGF separated from TCGF by only a few fractions. Co-elution of these two T cell-

derived lymphokines, having close m.w. and similar biochemical characteristics, previously has complicated the issue of lymphokine specificity. Furthermore, with the recent demonstration of the apparent B cell prolifer- ative effect of recombinant and affinity-purified IL 2 and of the existence of Tac receptor on a subset of activated B cells, IL 2-contaminated BCGF preparations render the specific role of BCGF in B cell activation and prolifer¬ ation controversial. Therefore, true separation of the co-eluting lymphokines has important implications.

The presently disclosed purification scheme involves multiple procedures, yet incorporates a single step for the removal of TCGF by hydroxylapatite chromatography. The remainder of the procedure after this separation step was analytically directed toward eliminating other contaminating proteins, an anion-exchange (DEAE-HPLC) and gel filtration on HPLC were used. The complete procedure yielded purification of BCGF with a specific activity of 10 ⁵ to 10 ⁶ U/mg of protein.

c. Preparation of a BCGF-Dependant Cell Line

For the preparation of B cells which are BCGF- dependant for proliferation, B cells are first isolated from human peripheral blood. B-cells are isolated essentially by negayive selection and are derived from the non-adherant E-rosette negative population of cells. Lymphocytes isolated in this manner are usually greater than are equal to 85% slg positive containing less than 2% E-rosette positive cells.

Following the preparation of B-cells from human peripheral blood, the isolated B-cells are activated with known B cell activators such as dextran sulfate staph protein A or anti-immunoglobulin, mu-chain specific. Such

activation procedures are well known to those of skill in the art.

Finally, the activated B cells are grown in the presence of highly purified BCGF (free of TCGF activity) for approximately one month to select for BCGF-dependant B cells. T lymphocytes, sometimes found to grow out during the initial phases of culture are routinely removed by E- rosette dependant procedures. The medium containing BCGF is replenished every 48-72 hours dependant upon cell growth. After approximately 30 days culture, if cells are proliferating, the cultures are maintained on BCGF and cell proliferation is continually monitored in terms of density and surface marker profile.

d. Factor Microassay

i. BCGF

In a typical experiment 0.2 x 10 BCGF dependent human B cells (88% surface Ig ,, <1% T,,+, and <4% NSE ) were cultured in the presence of anti-IgM (15 ug/ml; Bio- Rad) with samples or aliquots of the suspected BCGF factor at varying concentrations, in a final volume of 0.2 ml at 37°C in a 5% CO- incubator. Tritiated thymidine ( H-Tdr)

(6 Ci/mM, 1 uCi/0.2 ml) was added in the final 16 hr of a total 96-hr incubation period. Thymidine incorporation into the B cell DNA was used to assess the proliferative activity of BCGF. For the BCGF assay based on long-term

4 cultured B cells, 1.0 x 10 cells were cultured in a final volume of 0.2 ml RPMI containing 2% heat-inactivated serum with and without BCGF sample preparations. H-Tdr was added during the final 16 hr of the total 42-hr incubation. Units of BCGF activity were calculated by probit analysis whereby one unit of activity is defined as the amount of BCGF that induces 50% of maximal thymidine

incorporation as determined against a standard BCGF preparation. The ^' protein content of the sample was determined by the Bio-Rad protein assay (Bio-Rad), using bovine serum albumin as a standard protein.

ii. TCGF

For determination of TCGF bioactivity, 30- to 45- day-old T cell blasts, grown in. the present of TCGF, were washed free of factor and were used in culture at a final concentration of 1 x 10 /0.2 ml of culture volume with and without growth factor at 37°C for 72 hr (6). ³ H-Tdr was added during the final 16 hr. and incorporation was measured after cell harvesting.

Enrichment For BCGF Specific Messenger RNA

To enrich for BCGF-specific mRNA, relative size determinations for BCGF mRNA were made by sizing total poly A mRNA in association with in, vitro translation. Normal human T cells (500x10 ), in the presence of an obligate number of monocytes (5%) were stimulated with PHA (o.75%) for 20 hr at a density of 2 x 10 cells/mo. Total RNA was isolated using the RNase inhibior guanidinium isothiocyanate (cite), and CsCl gradient centrifugation. Poly A mRNA was obtained by one cycle elution of total RNA through an oligo-dT cellulose column. 300 ug of total RNA routinely contributed 10 ug of poly A RNA. Poly A mRNA (60 ug) was size fractionated on a 1% low melting agarose gel in the presence of 10 mM methyl-mercury hydroxide (cite); the gel was subsequently sliced. The RNA was eluted from each sliced gel fraction and injected into Xenopus laevis oocytes. Supernatants conditioned by the injected oocytes were assayed for BCGF and TCGF activity at multiple dilutions.

Shown in Fig. 1A is a representative microtiter culture assay on long term cultured B and T cells. TCGF activity coded for by the 12 S mRNA served as an internal marker for size as well as for in vitro translation.

3 Results are presented m terms of H-thymidine incorpor¬ ation at a 2.5% final concentration of oocyte supernatant. Ribosomal RNA (28 S and 18 S) were used as standard markers. Identification of TCGF mRNA serve as an internal marker for the fidelity of in vitro translation.

As demonstrated by Figure 1A, the majority of the TCGF activity is found in those mRNA fractions centering around 12 S. Interestingly, 16-18 S mRNA species are found to be responsible for the majority of the BCGF activity (Fig. 1 A). This permits the pooling of BCGF specific mRNA for subsequent gene cloning using a preparation virtually devoid of TCGF mRNA contamination.

f. Preparation of BCGF cDNA-Enriched Recombinant Vectors

Two micrograms of 16-18 S mRNA was used as a template to synthesize double stranded cDNA using reverse transcriptase for insertion into a Pst I digested and oligo-dT tailed pUC 9 vector essentially as described by Heideker et al. (1983), supra. The pUC 9 vector was selected with the prior assumption that the promoter of the B-galactosidase gene would direct the efficient expression of partial or full length cDNA molecules inserted in the correct sense orientation. Recombinant proteins would thus be translated as a fused product to a portion of the amino terminus of the B-galactosidase protein. This presumes that the cDNA molecules have to be placed in an open reading frame with respect to the aminoterminal-coding end of the B-galactosidase gene. Initial screening of the resultant E. coli library

revealed approximately 700 clones which contained cDNA inserts.

It was hoped that fusion proteins created between the aminoterminal end of B-galactosidase and peptides encoded by BCGF cDNA molecules might be biologically active for BCGF activity. cDNA containing colonies were therefore grown and pooled into groups of 17 colonies each. 10 mis of bacteria in each pool were grown and cell extracts were made by sonication of cells followed by high speed centrifugation. Crude cell extracts containing 0.2 mM PMSF were then dialyzed against RPMI media. The dialzyed cell extracts were screened for BCGF activity using the microassay and the data demonstrated that only two pools consistently produced the lymphokine activity when compared to extracts from bacteria containing only the parental vector pUC 9. Individual colonies from these two pools were grown and screened for BCGF bioactivity. As shown in Table 1, each pool here found to contain only a single colony (the colony # denoted in parentheses) which exhibited BCGF activity. Plasmid DNA's were isolated and characterized for size. cDNA inserts in plasmids pARJ 45 (pool 4) and pARJ 43 (pool 11) were of approximately 400 bp and 700 bp respectively.

TABLE 1 .

BIOLOGICAL ACTIVITY ASSAY FOR DETECTION OF BCGF cDNA CLONES.

cDNA Pools cpm [ H]-Tdr incorporation

(Colony) E. coli extract (%)

5 10 20

pUC9 202 296 504

4(45) 916 3934 7808

11(43) 222 2136 2296

5(55) 326 297 399

17(36) 0 5 0

Control BCGF 124 42288 15638 16420

10 ml of E___ coli containing various cDNA inserts were grown to OD, _nn = 0.3 and then induced in the presence of ImM isopropyl-B-D-galactopyranoside and 0.004% 5-bromo-4- chloro-3-inidolyl-B-D-galactopyranoside for next 3 hr. Bacteria were harvested and cell extracts were prepared by sonication. Cells alone incorporated 1220 cpm and the results reported above represents cpm/well minus the incorporation of cells alone.

EXAMPLE I: PARJ45 and PARJ43

To demonstrate that plasmids pARJ 45 and pARJ 43 contained BCGF specific cDNA inserts, these plasmids were used to hybrid select BCGF specific mRNA as described by Parnes et al. (1981), Proc. Natl. Acad, Sci. U.S.A., 78: 2253. This assay serves to identify clones bearing BCGF- coding DNA sequences by the ability of the clonal DNA to select BCGF mRNA by DNA/RNA hybridization. Following such "hybrid selection" of mRNA, the mRNA is analyzed for its ability to prime the translation of proteins exhibiting BCGF activity. In particular, DNA was prepared from colonies containing these plasmids and was bound to nitrocellulose filters. The filters were hybridized to mRNA from activated T cells. As a negative control, pUC 9 and two other insert-containing plasmids (pARJ 36 and pARJ 55; inserts not related to the BCGF gene) were analyzed. mRNA that hybridized to the various plasmid DNAs was injected into Xenopus oocytes and secreted products were assayed for BCGF activity.

Figure IB shows that plasmids pARJ 45 and pARJ 43 specifically hybridized to BCGF mRNA. Briefly, plasmid

DNAs (2 ug) were isolated and hybridized to poly A RNA (400 ug) from activated T cells as described by Fung et al. (1984) Nature, 307: 233. Bound mRNA was eluted by boiling for 1 minute and microinjected in Xenopus oocytes.

Secreted products were assayed on long term cultured human

B cells at multiple concentrations. Results are repre-

3 sented in Fig. IB in terms of H-Tdr incorporation in 12 x

10 long term cultured B cells. Repeated experiments showed that both these plasmids were equally potent in the positive selection of BCGF mRNA. It should be emphasized the pUC 9 and the non-BCGF containing plasmids were not capable of specifically selecting BCGF mRNA.

Example II: Human BCGF Expression

To demonstrate the specificity of BCGF expressed in bacteria, 100 ml cultures of bacteria containing either pARJ 45, pARJ 43 or the parental plasmid pUC 9 were grown to an optical density of 1.0 at 600 nm. The bacteria were pelleted and washed once with hypotonic buffer containing 20 mH Hepes (pH 7.5), 1.5 mM Mg (OAc)_, 3.6 mM CaCl ₂ , and 2 mM 2-mercaptoethanol. Cell pellets were resuspended in 5 ml of hypotonic buffer and sonicated three times at 30 sec intervals by immersing the tubes in ice chilled water. Cell extracts (supernatants) were obtained by centrifuga¬ tion at 25000 rpm for 30 min at 4 C and were immediately dialyzed overnight against 20 mM NaP0 ₄ (pH 7.5) buffer containing 0.2 mM PMSF. Dialyzed cell extracts were loaded onto a Sephadex G-50 gel filtration column. Proteins were eluted with 20 mM phosphate buffer (pH 7.0) and fractions were assayed for the BCGF activity using long term cultured B cells at various concentrations.

Occassionally nonspecific bioactivity and/or cyto- toxic effects are noted in response to crude cell extracts irrespective of their source. These complications can generally be avoided by limited sonication of cells, high speed centrifugation and extensive dialysis in addition to chromatography.

As seen in Fig. 2 panel A, the extract from bacteria containing pUC 9 was negative for BCGF activity, whereas BCGF activity was found in the extracts from bacteria harboring pARJ 45 and pARJ 43. It should be emphasized that the BCGF containing biologically active extracts may represent truncated proteins of approximately 14kD (pARJ 43) and 8-10 kD (pARJ 45). The size estimations were made by direct comparison of the cell extracts with mature BCGF calibrated on the same gel filtration column. The size

estimate with and respect to pARJ 43 was confirmed by DNA sequence techniques (see Example II).

Specificity of the human BCGF activity in the bacterial extracts may also be demonstrated in a conven¬ tional BCGF assay system based upon comitogenesis of anti-u activated B cells as described in the Maizel et al. (1983), P.N.A.S. , 80: 5047. As seen in Fig. 2, panel B, those extracts from plasmid pARJ 43, which possessed BCGF activity on long term cultured B cells also possessed activity in the anti-u based assay. No activity was detected in the fractions derived from pUC 9. Fractions from pARJ 45 were also found active. These results indicate that recombinant BCGF is capable of exhibitng the same biological activities as natural BCGF produced by human lymphocytes.

Plasmid pARJ 43 was also tested by northern analysis for its specifi ity of interaction with BCGF mRNA using total mRNA from PHA activated T cells. Northern blot anaysis was carried out by electrophoresis of 8 ug and 24 ug of poly A mRNA from PHA activated T cells (20 hr) on a 1% glyoxal-agarose gel. RNA samples were denatured with glyoxal for 1 hr at 50°C before gel electrophoresis. After elecrophoresis, the RNA was transferred to a nitro¬ cellulose paper and the blots were baked for 3 hrs at

80°C. The blots were prehybridized and hybridized with

32 P-labelled BCGF cDNA (pARJ 43) probe in the presence of

5 x SSC. 23 S and 16 S E_ _ι . coli ribosomal RNA were run in a parallel lane and visualized by ethidium bromide staining.

BCGF cDNA specifically hybridized to a 17 S mRNA species confirming our earlier prediction of mRNA size leading to the use of 16-18S mRNA for cDNA library construction. High molecular weight RNA species can be

attributed to the presence of nuclear precursors. However, when three fold more mRNA was used, the BCGF probe also detected a 11 s mRNA species . This northern analysis was done under low-stringency conditions (5 x o standard saline citrate, 42 C) . Furthermore, northern analysis using mRNA from a cell line devoid of BCGF secretory activity (i.e. monocytes) failed to show any hybridization with pARJ 43.

EXAMPLE III: RESTRICTION MAP AND SEQUENCE CHARACTERIZATION

Figure 3 depicts the restriction map of recombinant clone pARJ 43. The DNA segment insert of pARJ 43 which bears the sequence coding for BCGF activity is slightly greater than 700 base pairs in length and is transcribed in the indicated direction.

The coding sequence of the BCGF, as determined by standard DNA sequencing techniques, is presented in Figure 4. The corresponding BCGF amino acid sequence shown therein was derived from a consideration of the start, stop and amino-acid specifying codons found in this particular BCGF sequence. It will be appreciated that the DNA and amino acid sequences which precede the zero point (designated by negative numbers above the sequence segment) are derived from the B-galactosidase structural gene.

It will be appreciated that from a consideration of triplet codon degeneracy, that many alterations and changes may be made in the DNA sequence displayed in Figure 4 without departing from the spirit and scope of the present invention. That is, numerous changes may be made in the DNA sequence consistent with our knowledge of

code degeneracy and still obtain a DNA sequence which codes for the corresponding BCGF amino acid sequence of Figure 4.

Moreover, it is believed that many changes may be made in the amino acid sequence displayed in Figure 4 and still obtain a protein which exhibits BCGF activity. For example, it has been found by Kyte et al. (1982), J. Mol. Biol. , 157 :105, that certain amino acids may be substituted for other amino acids having a similar hydropathic index, and still retain the biologic activity of the protein. As displayed in Table 2 below, amino acids are assigned a hydropathic index on the basis of their hydrophobicity and charge characteristics. It is believed that the relative hydropathic character of the amino acid determines the secondary structure of the resultant protein, which in turn defines the interaction of the protein with its receptor. In the case of BCGF, it is believed that biological functional equivalents of BCGF may be obtained by substitution of amino acids having similar hydropathic values. As used herein, a biological functional equivalent of BCGF is defined as a protein that is functionally equivalent to BCGF in terms of biological activity. Thus, for example, isoleucine, which has a hydropathic index of +4.5, can be substituted for valine (+4.2) or leucine (+3.8), and still obtain a protein having like biological activity. Alternatively, at the other end of the scale, lysine (-3.9) can be substituted for arginine (-4.5), and so on. In general, it is believed that amino acids can be successfully substituted where such amino acid has a hydropathic score of within about +/- 1 hydropathic index unit of the replaced amino acid.

Table 2

Amino Acid Hydropathic Index

Isoleucine 4.5

Valine 4.2

Leucine 3.8

Phenylalanine 2.8

Cysteine/cystine 2.5

Methionine 1.9

Alanine 1.8

Glycine -0-4

Threonine -0.7

Tryptophan -0.9

Serine -0.8

Tyrosine -1.3

Proline -1.6

Histidine -3.2

Glutamic Acid -3.5

Glutamine -3.5

Aspartic Acid -3.5

Asparagine -3.5

Lysine -3.9

Arginine -4.5

g. Deposit of Microorganisms

E. coli SB229 (rec A ^~ , JM 103, Tet ) bearing the preferred recombinant plasmid pARJ43 was deposited with the American Type Culture Collection, 12301 Parklawn Drive, Rockville, Maryland on April 18, 1986 in accordance with the Budapest Treaty and assigned ATCC accession number 67099.

h. Purification of Recombinant BCGF

For clinical applications in particular, it will generally be desirable to obtain recombinant derived BCGF

• 5 in a relatively purified form. In general, purification of proteins from recombinant sources is well known in the art. Typically, the recombinant cells are cultured in an acceptable bacterial culture medium or broth under conditions which allow for the transcription and

10 subsequent translation of plasmid borne genes. Where the recombinant gene is associated with control sequences of a known promoter, such as the lac Z promoter found in the pUC-9 vector and numerous other vectors, it will be desirable to include in the cell culture mixture an agent

15 to selectively stimulate the transcription of such promoter, such as IPTG in the case of genes under the control of the lac Z promoter. Other transcription stimulatory agents as are known in the art may be included depending on the particular promoter employed.

20

After the recombinant cells have been cultured to a selected density, for example, an optical density of about 1.0 at 600 nm has been found to provide sufficient quantities of the recombinant BCGF protein, the cells are

25 typically hypotonically shocked to permeabilize the cells and thereby facilitate the obtaining of the protein from the cell culture mixture. Where the recombinant gene is constructed to include a "leader sequence" on the resultant recombinant protein, the cells are then removed

30 from the culture medium and the supernatant retained. As is known in the art, the inclusion of a leader sequence on the amino terminus of the protein directs the secretion of the protein by the recombinant cell, thereby obviating a need to disrupt the cell to obtain the recombinant

35 protein. Thus, for embodiments which include a leader sequence, the cells are hypotonically shocked, the intact

cells removed by centrifugation or other similar means and the supernatant retained as a starting source for protein purification.

However, as in the case of pARJ 43 and pARJ 45 where there is no leader sequence, and subsequently no large amount of secretion of the recombinant protein by the recombinant cells, it will generally be desirable to first remove the cells from the culture medium, wash and hypotonically shock the cells, and then disrupt the cells, for example, by sonication (see Example II). The cellular debris is then removed and the supernatant retained for BCGF protein purification.

Thus, the recombinant BCGF cell culture mixture, defined herein as any culture mixture which includes a recombinant microorganism producing recombinant BCGF, or extracts, lysates or sonicates of such microorganisms, is used for further purification of recombinant derived BCGF. The supernatants are rendered devoid of insoluble cellular material and subjected to molecular fractionation to obtain a fraction which includes the recombinant BCGF in a relatively purified form. As used herein, relatively purified recombinant BCGF is defined as recombinant BCGF that is relatively devoid of contaminating biological or immunological activities. Numerous molecular fractionation techniques may be devised by those of skill in the art to provide a relatively purified fraction, however, the classical scheme for the relative purification of natural BCGF from conditioned media

(described above in section b.) will work well in this regard.

Alternatively, one may desire to synthesize the BCGF protein by one of the various protein or DNA synthesizing techniques well known to those of skill in the art, from a

consideration of the amino acid or nucleic acid sequence information provided by the present disclosure. Thus, one may alternatively desire to fabricate a synthetic gene from a consideration of the the supplied DNA sequence or a synthetic protein from the amino acid sequence. Both approaches will incorporate the advantages of the present invention by allowing the production of a BCGF preparation that will be devoid of contaminating cytokine and related activities. For simplicity sake, all such synthetically derived species are intended to be included by reference to recombinant derived BCGF.

Following the obtaining of a relatively purified BCGF fraction, the recombinant molecule may be formulated into a pharmaceutical composition for clinical application. Typically, the composition is formulated to include effective concentrations of the BCGF protein together with an acceptable pharmaceutical diluent or excipient. For example, the compositions typically include various salts, buffers and/or stabilizing and preservative agents. For parenteral administration, solutions of the novel compound may include sesame or peanut oil, aqueous propylene glycol, or the like, together in an aqueous sterile solution. Such aqueous solutions should be suitably buffered if necessary and the liquid diluent first rendered isotonic with sufficient saline or glucose. Such solutions are particularly suitable for intravenous, intramuscular, subcutaneous and intraperitoneal administration. In this regard, the sterile aqueous media employed are all readily available by standard techniques known to those skilled in the art.

Effective concentrations of recombinant BCGF will generally be determined based on the particular clinical application envisioned. However, treatment regimens employing BCGF typically require from about 10 5 to 107

units BCGF per application. Therefore, it will typically be desirable to provide formulations which include BCGF in concentrations sufficient to allow the administration of such dose units by whatever route is employed.

Human BCGF as described in the present application has been shown to be involved in the clonal expansion of normal mature human b-cells. This effector function will therefore likely possess utility both in those clinical situations of idiopathic immunodeficiencies and in those clinical situations where a B cell immunodeficiency state is induced by pharmocologic agents. In addition, this growth factor may play a role in further understanding the B cell receptor for the growth factor. The receptor has recently been demonstrated to be present on several neoplasms of B cell origin. Therefore, the described growth factor may help our understanding of receptor binding sites and thereby provide a means for the development of anti-receptor reagents.

* * *

Although the foregoing invention has been described in terms of preferred embodiments, those of skill in the art will recognize that various modifications may be undertaken without departing from the scope of the appended claims. For example, numerous alternatives and modifications are known for the various genetic and enzymatic manipulations described herein will be apparent to skilled molecular biologists with the assistance and in light of the present disclosure. The present inventors consider such modifications to be within the scope of the present invention. In addition, although the invention is described in terms of a prokaryotic host/vector expression system, there is no reason why human BCGF gene isolated as described herein cannot be expressed in one of the several eukaryotic host/vector systems known in the art. Accord-

ingly, these and all other modifications are considered to be within the scope of the appended claims.